Perioperative stress dose steroids: do they make a difference?

Abstract

Since the Nobel prize-winning description of the use of synthetic glucocorticoids in the treatment of rheumatoid arthritis by Hench and colleagues in 1949, steroid use has become ubiquitous among patients with inflammatory conditions. Although the benefits of steroids are clear for these patients, it has become apparent that there are very serious potential side effects of the treatment. The potential dangers of chronic steroid use in surgical patients were first illustrated by Fraser and colleagues in 1952. They describe a steroid-dependent patient undergoing major orthopaedic surgery who died of intractable hypotension postoperatively. An autopsy revealed adrenal atrophy, and the authors conclude that the patient died from adrenocortical insufficiency secondary to adrenal suppression. Since the publication of this singlecase report, supraphysiologic “stress dose” steroids have become a regular part of the perioperative management of patients treated longterm with steroids. In individuals with an intact hypothalamic-pituitaryadrenal (HPA) axis undergoing the stress of an operation, the adrenal glands increase their baseline secretion of cortisol. Although the details are not clear, it is thought that increased circulating cortisol helps maintain hemodynamic stability in the face of stress. Patients treated longterm with exogenous steroids have a blunted HPA axis response to surgical stress when compared with normal controls, resulting in lower levels of circulating cortisol. Based on this physiologic data and a handful of case reports like that of Fraser and colleagues, several reviews have been published over the past 50 years recommending supraphysiologic corticosteroid regimens for such patients undergoing operations. Although it is known that endogenous steroid levels become elevated during surgical stress, it is uncertain whether a reduced level of circulating cortisol results in hypotensive crisis. Kehlet and Binder followed 31 patients treated with exogenous steroids through major surgical procedures. Each patient was shown to have HPA axis suppression by a failed response to an adrenocorticotropic hormone (ACTH) stimulation test. None of these patients was treated with steroids perioperatively, and none of the patients showed any symptoms of acute adrenocortical insufficiency. Nine of these patients had subnormal plasma cortisol levels throughout the operation. Although surrogate biochemical markers of adrenal function were shown to be suppressed, there was no correlation between these data and patients’ clinical outcomes when subjected to the stress of a surgical procedure. In 1976, Kehlet reviewed the world literature and found 57 case reports of perioperative hypotension attributed to adrenal insufficiency. His investigation revealed that only three of these reports had sufficient clinical and biochemical data to support this diagnosis. Kehlet’s work began to undermine the tenets of perioperative stress dose steroids over 20 years ago. Studying an animal model, Udelsman and colleagues have also challenged the necessity of stress dose steroids in the perioperative period. First, they surgically removed the adrenal glands of a cohort of primates. They then randomized the monkeys to receive supraphysiologic (10 times normal dose), physiologic, and subphysiologic (one-tenth normal dose) perioperative steroid coverage for an open cholecystectomy. The animals treated with subphysiologic doses of steroids demonstrated a markedly decreased mean blood pressure, increased incidence of electrolyte abnormalities, and increased mortality when compared with normal controls, but there was no difference in these outcomes when the monkeys treated with physiologic doses of steroids were compared with monkeys with supraphysiologic treatment or nonadrenalectomized controls. Udelsman and No competing interests declared.